Hashlogics
Glossary

What is the work product doctrine?

Two people can write the same memo about the same case, and only one version is protected. The difference is whether a lawyer asked for it.

Work product doctrine

Attorney work productWork product privilege

Work product is material a lawyer prepares for expected litigation, or has someone prepare. The other side cannot get that analysis in discovery.

People treat this as a twin of attorney-client privilege. They are different protections with different tests. Privilege covers a confidential communication between lawyer and client seeking legal advice. Work product covers the material itself. It asks two questions. Was this made because litigation was expected? And did a lawyer ask for it?

That second question is where software now lives. Anticipation of litigation is usually easy to establish. Direction of counsel is a fact about a workflow, and most systems never record it.

Why it matters

A 2026 ruling made direction a data field

In United States v. Heppner was decided in the Southern District of New York in February 2026. A defendant used a consumer AI app to research his own case and build strategy documents. Judge Jed Rakoff held they were not work product. They had not been prepared by counsel or at counsel's behest, and they did not reflect defense counsel's strategy. Anticipation of litigation was not enough on its own.

Read that as a specification. Any product where a client, paralegal or agent generates analysis has to capture three things. That a lawyer asked for the task, which lawyer, and for which matter. Not in a comment box. In the record, attached to the artefact, at the moment it is created. Adding it later is guesswork. The court also held that forwarding a document to a lawyer does not cloak it afterwards.

The two-part testLive
  1. TriggerLitigation reasonably expected
  2. DirectionCounsel asked for the task
  3. CreationAnalysis, memo, model output
  4. RecordWho directed it, and when
  5. AssertLogged on the privilege log

Skip the fourth node and the first three become an argument rather than a fact. That is the gap Heppner turned into a holding.

Questions, answered

Questions people ask next

01How is work product different from attorney-client privilege?

Privilege protects a confidential communication between lawyer and client for legal advice; work product protects material prepared because litigation was expected. A document can be one, both or neither. In Heppner the AI-generated documents failed both tests, but for different reasons: no confidentiality for privilege, no direction of counsel for work product.

02Is work product protection absolute?

No. Ordinary work product can be obtained by an opponent who shows substantial need and an inability to get the equivalent elsewhere. Material capturing a lawyer's thinking and legal theories gets stronger protection. That split is a good reason to keep analysis separate from raw facts in your data model.

03Does a client's own research count as work product?

Usually not, on the Heppner reasoning. A client researching their own case is not counsel and is not acting at counsel's direction, so the output reflects the client's thinking rather than the lawyer's strategy. The court left open that a properly structured arrangement where counsel directs a consultant's work can be different.

Verified
Start

Anyone can ship the agent. We answer the pager.

We build AI agents and automation, then stay on under an agreed service level. A senior engineer reads every brief, and your call gets scheduled within 24 hours.

What happens next

  1. 01

    You send a brief or book a call

    Two minutes, whichever you prefer.

  2. 02

    A senior engineer replies within 24 hours

    Not a sales rep.

  3. 03

    Honest scoping, in writing

    And if we’re not the right fit, we say so.

Abdul Basit, CEO of Hashlogics

“I started Hashlogics because too many teams ship a demo, get paid, and disappear. We build to a standard we’d run ourselves — and we stay to keep it running.”

Abdul Basit · CEO · a direct line

Not ready to talk? Take the checklist.

12 questions to ask any AI agency before you sign. They separate a demo shop from a team that ships to production.

Get the checklist

Free · no newsletter